bool Parser::ParseComment(token& out, bool isBlockComment)
{
line = cur_line;
column = GetColumn();
text.str += 2; // skip over the comment start
if (isBlockComment) {
// scan until */
for ( ; text.str < text.end; ++text.str) {
char ch = *text.str;
if (ch == '\r' || ch == '\n') {
SkipNewLine(ch);
continue;
} else if (ch == '*' && text.str[1] == '/') {
out.end = text.str += 2;
return true;
}
}
} else {
// scan until \r\n or \n
for ( ; text.str < text.end; ++text.str) {
char ch = *text.str;
if (ch == '\r' || ch == '\n') {
out.end = text.str;
SkipNewLine(ch);
++text.str; // skip \n
return true; // done
}
}
}
return true;
}
void LanguagePack::ParseLine(IStringReader *reader)
{
SkipWhitespace(reader);
codepoint_t codepoint;
if (reader->TryPeek(&codepoint))
{
switch (codepoint) {
case '#':
SkipToEndOfLine(reader);
break;
case '[':
ParseGroupObject(reader);
break;
case '<':
ParseGroupScenario(reader);
break;
case '\r':
case '\n':
break;
default:
ParseString(reader);
break;
}
SkipToEndOfLine(reader);
SkipNewLine(reader);
}
}
TRISTATE CTextParser::ReadLine(CTextParser* pcLine)
{
char* szStart;
char* szEnd;
pcLine->Init();
ReturnErrorOnErrorAndFalse(FindStartOfLine());
szStart = mszParserPos;
ReturnErrorOnErrorAndFalse(FindEndOfLine());
szEnd = mszParserPos;
SkipNewLine();
if (mbOutsideText)
{
szEnd++;
}
pcLine->Init(szStart, (int)(szEnd - szStart));
pcLine->mbAnnotated = mbAnnotated;
return TRITRUE;
}
bool Parser::GetToken(token& out)
{
// trim leading whitespace first:
for (; text.str < text.end; ++text.str)
{
char ch = *text.str;
if (ch == ' ' || ch == '\t') continue;
if (ch == '\r' || ch == '\n') SkipNewLine(ch);
else break; // not ' ' or '\t', so finish trim
}
out.end = out.str = text.str; // mark down start and default end
column = GetColumn();
for ( ; text.str < text.end; ++text.str)
{
char ch = *text.str;
if (ch == '\r' || ch == '\n') {
SkipNewLine(ch);
continue;
}
if (ch == '/') { // possible comment start?
char c2 = text.str[1];
if (c2 == '/' || c2 == '*') // line comment '//' or block comment '/*'
return ParseComment(out, c2 == '*'); // parse the comment
}
else if (ch == '"')
return ParseStringLiteral(out); // skip the literal
else if (ch == '\'')
return ParseCharLiteral(out);
if (BreakerIndex[ch] == 1) // token starts with a breaker char
{
line = cur_line;
column = GetColumn();
char c2 = *++text.str;
char c3 = text.str[1];
if (c3 == '=' && (
(ch == '<' && c2 == '<')|| // <<=
(ch == '>' && c2 == '>'))) // >>=
{
text.str += 2; // 3 char token parsed
}
else if ((c2 == '=' && (
ch == '<' || // <=
ch == '>' || // >=
ch == '=' || // ==
ch == '!' || // !=
ch == '+' || // +=
ch == '-' || // -=
ch == '*' || // *=
ch == '/' || // /=
ch == '%' || // %=
ch == '&' || // &=
ch == '|' || // |=
ch == '^' // ^=
)) ||
(ch == '+' && c2 == '+') || // ++
(ch == '-' && c2 == '-') || // --
(ch == '-' && c2 == '>') || // ->
(ch == '?' && c2 == '?') || // ??
(ch == '&' && c2 == '&') || // &&
(ch == '|' && c2 == '|') || // ||
(ch == '<' && c2 == '<') || // <<
(ch == '>' && c2 == '>') || // >>
(ch == ':' && c2 == ':')) // ::
{
++text.str; // 2 char token parsed
}
out.end = text.str;
return true; // 1 char token parsed
}
// so this is a literal or a statement?
return ParseStatement(out);
}
return false; // end of buffer
}
